N.A. Abu Osman et al. (Eds.): BIOMED 2011, IFMBE Proceedings 35, pp. 801–804, 2011. www.springerlink.com The Development of Muscle Training System Using the Electromyogram and Interactive Game for Physical Rehabilitation K.S. Kim 1 , J.H. Kang 1 , Y.H. Lee 1 , C.S. Moon 2 , H.H. Choi 1 , and C.W. Mun 1,3 1 Inje University/Biomedical Engineering, Obang-dong, Gimhae, Gyoungnam, 621-749, Republic of Korea 2 KMG Ltd, Saha-gu, Busan, Republic of Korea 3 Inje University/UHRC, Obang-dong, Gimhae, Gyoungnam, 621-749, Republic of Korea Abstract— In this study, muscle training system was devel- oped for effective rehabilitation treatment. This system con- sists of the electromyogram (EMG) acquisition module, the graphic user interface (GUI) system, and game contents of each channel mode. EMG acquisition module acquires EMG signals and transmits to GUI. GUI performs to control the direction in for game operation using acquired EMG signals. The obtained data were analyzed using laboratory-developed software based on Microsoft Foundation Class (MFC) of Vis- ual Studio 2008. The games were developed as each-channel mode for various exercises in rehabilitation. All the games were implemented in Flash CS4 using Action Script 2.0. This system was to overcome the boredom and monotony in reha- bilitation. We proposed the method which improves the pa- tient's participation willingness and motivates them in reha- bilitation. Keywords— Electromyogram, Interactive game, Rehabilita- tion. I. INTRODUCTION Many studies on electromyogram (EMG) systems have been reported in several fields, such as in the fields provid- ing assistance for disabled and for rehabilitation training, and in the field of device control [1-3]. Nevertheless, an EMG device has been passively applied to a rehabilitation treatment and has some limitations in terms of the patient’s involvment. It is well known that the patient should actively participate in the treatment to improve this limitations. The EMG system combined with the game would help the pa- tients actively participate in rehabilitation training. So the development of interesting training system is required for effective rehabilitation [4,5]. This kind of interface method is called the interactive game. The interactive game inter- face is a user interface in which a person interacts with digital information through the physical environment [6]. The EMG can generate necessary control commands for the game system result from the muscular movements. And it can be used as a method of interactive game interface [7,8]. The recognition of exercise directions is required to control device or game interface for rehabilitation. In the previous study, we proposed a reliable algorithm without time delay to recognize the defined wrist-motion patterns [7,9]. Rehabilitation and physical therapy are essential elements for patients to recover from functional disabilities. Because rehabilitation requires strenuous and repeated movement, it can be helpful to add fun and enjoyment using game device like Nintendo wii and play station [10~12]. If using EMG signal compared with motion recognition through an accel- eration sensor, variety information about the muscles as well as the motion recognition can be obtained. Generally game contents have been made for the public, but for the patient, it may be somewhat difficult to use. Thus, the ne- cessity of the customized game contents for rehabilitation has been being increased. In this study, we proposed the multi-channel EMG mod- ule combined with customized game contents for effective rehabilitation which improves the patient's participation willingness and motivate them in rehabilitation by overcom- ing the boredom and monotony in rehabilitation. II. METHODS The proposed training system consists of the EMG ac- quisition module, the graphic user interface and game con- tents for each channel mode. A. EMG Acquisition Module The EMG acquisition module was configured to acquire up to 4 channel signals simultaneously. The semi permanent dry type electrodes (KOREC, Korea) as shown in Fig. 1 were used as EMG sensor [13]. This dry sensor includes differen- tial amplifier, bans-pass filter (10~500Hz) and band reject filter (60Hz). The valid EMG signal was acquired by amplifi- cation and filtering of frequency band without separating analog circuit from dry sensor. The obtained signal received from designed EMG module was transferred to PC after signal processing. The sampling frequency of the 12-bit A/D converter was 1 kHz. Self-developed program was used to operate ATMEGA128 (Atmel, Inc, USA), and a converted signal was transmitted to software through communication